A Rasch Analysis of a Measure of Graph Selection and Reasoning for Dynamic Situations

• In Event: Rasch Measurement Innovations

Abstract

This study reports results from two Rasch modeling for a measure of Graph Selection and Reasoning for Dynamic Situations. Data was collected from 673 undergraduate students in college algebra across four institutions. Both models showed scales worked well to measure each construct and the Rasch models fit the data well. The Reasoning scale is specifically unique in that this is the first time students’ graph reasoning has been statistically supported as a hierarchical scale. As the reasoning scores began with qualitative coding of written responses, this study offers a validated way to code from a rubric to support a valid quantitative scale.

Authors

• Courtney Donovan, University of Colorado - Denver
  Presenting Author

• Heather Lynn Johnson, University of Colorado - Denver
  Presenting Author

• Livvia Bechtold, University of Colorado - Denver
  Presenting Author

• Robert Knurek
  Non-Presenting Author

• Kristin A Whitmore, University of Colorado - Denver
  Non-Presenting Author